3 k c h 196s
IRREVERSIBLE EXZYME INHIBITORS. CXV
method A a t -5 to 0” except E t & was used as the acid acceptor. Method C. N-(p-Aminomethylphenyl)-3-chlorophenoxyacetamide Hydrochloride (23).-A solution of 2.86 g (10 mmoles) of 24 and 1.5 g (15 mmoles) of EtS03H in 200 ml of 957, aqueous methoxyethanol was stirred with 0.10 g of decolorizing carbon, then filtered. After the addition of 50 mg of PtO,, the mixture was shaken with Hz at 2-3 atm. When about one-fourth of the calculated amount of H2 had been consumed, hydrogenation ceased. The mixture was re-treated with decolorizing carbon and filtered, and then 50 mg of fresh PtOz was added. After being shaken with HS at 2-3 atm for an additional 24 hr, hydrogenation was essentially complete. The filtered solution was diliited with 600 ml of HSO and a small amount of 24 was removed by filtration. The filtrate was adjusted to pH 11-13 with lOyo KaOH, then extracted with three 50-ml portions of CHC13. The combined extracts were dried (MgS04),then saturated with HCI gas. The product was collected on a filter and recrystallized from EtOH; yield, 1.97 g (607,) of white crystals, mp 253-256 dec. See Table I1 for additional data. Method D.-A mixture of 2.11 g (5 mmoles) of 32, 50 ml of lIeOEtOH, and 1.0 ml (26 mmoles) of 85% S2H4.H10 was reflrixed with stirring for 20 min when solution was complete. Solvent was removed by spin evaporation zn 2jacuo. The residue was stirred with 60 ml of 1 S HC1 a i about 85’ for 15 min; the insoluble phthalhydrazide containing some 32 was removed by filtration and washed with HzO. The phthalhydrazide contain-
245
ing some 32 was again treated with X S H for ~ 20 min &s above. The combined HCI solutions were brought to about pH 13 with 50% SaOH. The mixture was extracted with two 100-ml port,ioris of Et20. The combined extracts were dried (SlgSO4), then treated with HC1 gas until no more product separated. The product was collected and recrystallized from EtOH; yield 0.15 g ( l o % ) , mp 231-2s560 dec. Method E was the same as method C except no HzO or EtS03H was added t o the reduction mixture. Method F. N-(m-Bromoaeetamidobenzyl)-3,4-dichlorophenoxyacetamide (4).-To a solution of 323 mg (1 mmole) of 48 in 10 ml of l\fe2C0was added 390 mg (1.5 mmoles) of bromoacetic anhydride.” The mixtiire was stirred for 4 hr, then diluted (H20) if the product had not separated. The prodrict was collected on a filter, then washed (;\Ie2CO-H20, 5% S a H C 0 3 , HZO). Recrystallization from EtOH afforded 300 mg (67%) of white needles, mp 133-157’, that gave a negative Brat,ton-lIarshall test for arylamine and a positive 4-(p-nitrobenzyl)pyridinetest for activated halogen.” See Table I1 for additional data. Method G . 3-(m-Chlorophenoxyacetamidomethyl)-3’-fluorosulfonyldiphenylurea (13).-To a solution of 291 mg (1 mmole) of 48 in 13 ml of CHC13 was added 220 mg (1.1 mmoles) of mflriorosiilfoiiylpheiiyl isocyanate iii 1.5 ml of CHC13. After 20 min, the mixture was filtered and the product was washed v,-ith CHCI,. Recrystallization from EtOH gave 319 mg (63c0) of white crystals, mp 178-181”, which gave a negative Brattonllarshall test for arylamine.” See Table I1 for additional data.
Irreversible Enzyme Inhibitors. CXV. 1,2 Proteolytic Enzymes. V. Active- Site-Directed Irreversible Inhibitors of Trypsin Derived from p(Phenoxyal1roxy)benzamidines with a Terminal Sulfonyl Fluoride B. R. BAKERASD EDWARD H. ERICKSON Department of Chemistry, Unzcersitzl of Calzfomia at Santa Barbara, Santa Barbara, California
9S106
Received Sooewiber 30, 1967 Eight candidate irreversible inhibitors of trypsin have been synthesized from p-phenoxypropoxybenzamidine (3) and p-phenoxybutoxybenzamidine by insertion of a leaving group on the para position of the phenoxy moiety. The p-bromoacetamido derivative ( 6 ) of 3 was a good reversible inhibitor of trypsin, biit failed to show any irreversible inhibition. When the methylene bridge of 6 was increased to biityl, the resultant bromoacetamide (11) showed slow irreversible inhibition of trypsin with a half-life of about 5 hr. 31uch more dramatic results were obtained by insertion of a vi- or p-fluorosulfonylbenzamido or m-fluorosulfonylphenylureido group on the phenoxy moiety. Five irreversible inhibitors of this type could inactivate trypsin with a half-life of 3-6 min. The kinetics indicated that not only did these sulfonyl fluorides rapidly inactivate trypbin, but trypsin could catalyze the hydrolysis of the sulfonyl fluorides to the irreversibly ineffective sulfonic acids.
Selective inhibitors for serum proteases could have utility for chemotherapy of cardiovascular diseases and of organ tran~plantation.~Since a number of these enzymes are “tryptic” in character, our initial studies have concentrated on trypsin as a model. Furthermore, active-site-directed irreversible inhibitors4 of the exo type could be expected to have the necessary selectivity in inhibition of these closely related serum proteases. The design of such exo-type irreversible inhibitors is best performed stepwise: (a) the binding points of an inhibitor should be determined; (b) a position on the inhibitor should be found where large groups can be tolerated without interfering with formation of a reversible enzyme-inhibitor complex, a necessary inter(1) This work was generously supported b y Grant CA-08695 from t h e National Cancer Institute, U. S.Public Health Service. ( 2 ) For t h e previous paper see B. R. Baker a n d J. A. Hurlbut, J . M e d . Chem., 11, 241 (1968). (3) For a more detailed discussion of inhibition of these enzymes see €3. R. Baker and E. H. Erickson, abzd., 10, 1123 (1967), paper CVI of this series. (4) B. R . Baker, “Design of Active-Site-Directed Irreversible Inhibitors. T h e Organic Chemistry of the Enzymic Active-Site,” John W l e y and Sons, Inc., New York, N. Y . , 1967.
mediate in enzyme inactivation by the active-sitedirected mechanism; (e) a properly positioned leaving group should be placed on this large moiety which becomes juxtaposed to a nucleophilic group when complexed to the enzyme and has the proper electrophilicity to react with this enzymic nucleophile; (d) the irreversible inhibitor is further modified to give selective inhibition of closely related enzymes or isozymes. Since benzamidine is an excellent reversible inhibitor of trypsin,5 it was selected for further study in order to convert it to an active-site-directed irreversible inhibitor of the exo type. I n our first paper on t r y p ~ i n , ~ we established that phenoxyalkoxy groups could be substituted on benzamidine a t either the para (1) or ineta positions (2) with retention of binding; in fact
1, para isomers 2, neta isomers ( 5 ) 31. hfares-Guia and E. Shaw, J . B i d . Chem., 240, 1579 (1965).
"Iri
!i
2
4 I
1 li
2
li
II
roinpi~uiid~ of type 1 arid 2 \\-ere 2-&fold better inhibit (11's ( ) f t r y p i n than the parerit hizaniidiiie, t hii. meeting the bulk-tolerance requiremelit in b above. The I)lucement of suitable electrophilic groups oii tlic terminal phenoxy moiety of 1 that have given some new types of active-site-directed irrever4ble inhibitors of' t r y l ~ i is i the subject of this paper. Enzyme Results.- -p-(Pheiiosyl)rol)~l~~xy) berizamichic (3)3with a K , iiear 7 p&lJ (Table I) wa5 first selected for cvriveriioii to candidate irreversible inhibitor\*. 11Lt rotluctiori of the I,-brom":tcetaiiiido moiety gave' :I c~)nil)ourid(6) that was reversibly coniplexed to t r y ) sin slightly better thnri the 1)areiit 3. When trypsin i w 5 iricubatecl a t 37" with 17 pA1J 6, which is sufficient t o coiivert 79c/;. of the erizynie to the rate-determining sliecies [I.:I],fli i o irreversible irihibitioii occurred i i i 60 mi I 1. The discovery of the terniinal d f o ~ i yfluoride l group for inactivation of dihydrofolic reducta5e by suitably iuhtituted 4,G-diami1io-l,L>-dihydro-s-triazines~suggested that the terminal sulfonyl fluoride grouy) be attached to p - (phent ropyloxy) benzamidirie (3). The series to be invebtigated war the p-flr~olosnlfoiiylheiizaniidodrrivativc (4) i i f 3, 4 heirig ahoiit :i iwofold licttcr reversible iiihibitor thaii 3. * I t :i (,oiicciit~atiori of 4 (2.5 pJ1) yufficient to complex hSC/( ol t h r tryi)\iii, the eiizynie n : ~ h rapidly iiiactivated ; .ioc, iii:icti\wtioii occurred iii 3 iriiii :uid total inactivutioil I I I 1.i n i i i i . That :i reveiiible complex betwecii 4 : i i i d tryI)\*iiiI\ :I- a i l ohlig:itory iritcmiietfi:itr t o iiiacti! :I-
i l
March 196s
IRREVERSIBLE ExzYnrE INHIBITORS. CXV
247
hydrofolic reductase7 and chymotrypsin.2 In the latter case, this line curvature was proven to be due to a combination of inactivation of the enzyme and enzymecatalyzed hydrolysis of the sulfonyl fluoride by kinetic observations and by isolation of the corresponding sulfonic acid. Similar kinetics have now been observed with 4 and trypsin. That the sulfonyl fluoride (4) was not hydrolyzed by buffer was shown by preincubation of 4 at 37" in the absence of enzyme for 1 hr; when enzyme was added, the rate curve for inactivation mas the same as a standard without preincubation. The effect of p H on the relative rate of enzymecatalyzed hydrolysis of the sulfonyl fluoride and to the rate of inactivation of the enzyme by the sulfonyl fluoride was then determined. (a) At pH 8.4, 3.1 p M of 4 rapidly inactivated the enzyme; 50% inactivation occurred in 1 min and 8iyo in 15 min. This contrasts with pH 7.4 where 44y0 inactivation occurred in 15 min, then no further inactivation occurred. These results indicate that the inactivation is more rapid at pH 5.4 and the ratio of inactivation to hydrolysis is also increased. (b) At pH 6.5, 3.1 ItM of 4 inactivated the enzyme more slowly than a t pH 7.4; 30% inactivation occurred in 30 min and 4SY0 in 90 min. Since inactivation was still proceeding at 60 min, the sulfonyl fluoride was still present, in contrast to pH 7.4, where all of the sulfonyl fluoride was hydrolyzed in 15 min. Thus, both inactivation of the enzyme and hydrolysis of the sulfonyl fluoride are slower than a t pH 7.4. (c) At p H 5.0, no inactivation of the enzyme was detectable in 2 hr. Whether this lack of inactivation was due to failure of irreversible inhibition to occur or that enzyme-catalyzed hydrolysis was extremely rapid cannot be separated by the current methodology; the latter is unlikely in view of the slower irreversible inhibition at pH 6.5 than pH 7.4. These pH profile results indicate that both the rate of inactivation of the enzyme and the rate of enzyme-catalyzed hydrolysis are dependent upon the OH- concentration as previously observed with chymotrypsin;2 whether this increasing OH- concentration effects these reactions directly or effects ionization of enzymic groups cannot be stated. When the inactivated enzyme was treated with 10 mM mercaptoethanol at pH 7.4 no detectable regeneration of activity occurred in 10 min; whether or not other thiols such as thioacetate,*glutathione, mercaptoe t h ~ l a m i n eor , ~ thioglycollic acidlo could regenerate the activity by nucleophilic displacement of the covalently bound inhibitor deserves investigation, particularly in alkaline solution'O or S 171 urea.g Additional analogs of 4 were then synthesized for investigation (Table I). When the sulfonyl fluoride of 4 was moved to the meta position, the resultant 5 was about a twofold better reversible inhibitor; furthermore, 5 gave kinetics of irreversible inhibition similar to 4 when compared at similar [EI] concentrations. Five analogs (7-11) were then synthesized where the methylene bridge was increased from three to four car( 8 ) (a) K. E. Neet and D. E. Koshland, Jr.. Proc. Natl. Acad. Scz. c'. s., 56, 1606 (1966); (b) L. Polgar a n d XI. L. Bender, J. Am. Chem. Soc., 88, 3153 (1966).
bons. The p-bromoacetamide (11) at 14 p M , which gives S4% E1 complex, showed slow inactivation of trypsin with a half-life of about 5 hr; this result contrasts with the lower homolog (6) where no inactivation was seen at 17 p X (79% E1 complex). Three of the sulfonyl fluorides (7-9) in this series were both good reversible and irreversible inhibitors of trypsin with kinetic parameters similar to 4 and 5. However, when the chain of 7 was further extended by insertion of a paminobenzoyl residue, the resultant 10 was a 150-fold poorer reversible inhibitor than 7; this result indicates that the coplanar CEH,NHCOC6H&HCOC6H4 system of 10 cannot be tolerated within the enzyme-inhibitor complex. With the varying dimensions from the benzamidine moiety that complexes in the active site to the sulfonyl fluoride that attacks the enzyme, it is highly unlikely that the identical amino acid on trypsin is attacked by all of the sulfonyl fluorides in Table I. However, it is likely that a serine or threonine is attacked in each c a ~ e ; ~ *t9o prove such an attack, a study of the displacement of the sulfonate with mer~aptoethylamine~ or other mercaptans,lo followed by hydrolysis and identification of the S-substituted cysteine or S-substituted 0-mercapto-a-aminobutyric acid, would be worthwhile. Another problem results from the current study; can the ratio of enzyme inactivation by a sulfonyl fluoride to enzyme-catalyzed hydrolysis of this sulfonyl fluoride be changed by substitution on the phenylsulfonyl moiety? Such an approach on sulfonyl fluoride irreversible inhibitors of several enzymes is being pursued with some successes already being achieved with dihydrofolic reductase." The question of whether or not compounds of the type in Table I mill or will not irreversibly inhibit some of the serum "tryptic" enzymes3is also being pursued; if one of these sulfonyl fluorides inactivates two or more of these "tryptic" enzymes, can the irreversible inhibition be made selective by the bridge principle of specificity?12 The bridge principle of specificity is dependent upon covalent bond formation between the inhibitor and a nucleophilic group on the enzyme that is outside of the active site. Unfortunately, the size of the active site with a proteolytic enzyme using a protein as a substrate is difficult to ascertain since it is not yet known how many monomeric units of the substrate are in beneficial contact with the enzyme; therefore, it is also difficult to guess whether or not the sulfonyl fluorides in Table I attack exo or endo with respect to the active site. However, it mould not be too unreasonable to expect the active sites of proteolytic enzymes to differ at a distance from the catalytic part of the active site, particularly where the second or third amino acid from the bond cleavage may be held; if such differences that are still part of the active site do indeed exist, they should be detectable by appropriate irreversible inhibitors such as those in Table I. Chemistry.-Two routes to the key intermediates, p-(p-nitrophenoxya1koxy)benzamidines (16), were investigated which differed only in the order of reactions starting with p-cyanophenol (13) (see Scheme I). Alkylation of 13 with 3-bromopropyl p-nitrophenyl ether in DRIF in the presence of I(2CO8 at 70" afforded
(9) 8 . M. Gold a n d D. Fahrney, Bzochemzstry, 3, 783 (1964). (10) D. H. Strumeyer, W. N. White, and D. E. Koshland, Jr., f'roc. Natl. Acad. Scz. U. S.,60, 931 (1963).
(11) B. R. Baker and G. J. Lourens, unpublished. (12) See ref 4, pp 172-190, for a discussion of the bridge principle of specificity.
21 S
15
co
co
I
SOZF 9
4, imru isomer, n = 3 5, w t a isomer, n = 3 7, 1iii1.11isomer, n = 4 8, iriirriisomer,n=4
15 in 66% yield. The cyano group of 15 was converted t o the ttmidine (16a) via its imino ether hydrochloride i n 4374 yield for the two steps. Allternately,p-cyanoI)hericil (13) was converted t o p-hy(~roxybenxamidiiic (14) in 47% yield by the imino ether route,l'j which is found superior to the method of Partridge arid Short" irivolvirig fusion of 13 with SHISCS. Alkylatiori of 14 with the appropriate bromide in DMF-I I . IT. l'artridge and JV. F. Short. ibid., 300 (loti).
IRKEVERSIBLE ll;.uzuai~IXHIBITOHH. CXV
No.
R
%
Mil,
llettiod
lield
O C
I) 1) C 1) I) E 1) C A A
2.5 1 60C
rormula
Analyses
268-270 CjgHzsFS3OsS2 C, H, N >16.i dec J CzeH,,FK308Sz C, H, ?u' 15-167 6 C , ~ H , ~ B ~ N \ J ~ O ~ S C, H, N C ~ O H ~ O F N ~ O ~ S , C, I%,s 181~ 280-283 i 2Od 210-213 S 4 C ~ O H ~ O F K ~ O ~ S ~ C, H, S C ~ ~ H ~ I F S ~ O ~ S ~C, H, N 9 4 508 230-233 10 4 13' 294-296 C ~ ~ H ~ ~ N ~ O S S ? C, H, N C2,H28BrN30sS C, 11, x 348 198-200 11 4 CL~I%~~~JO# C, H,N 66 187-1928 16:t 1 39 223-228& l6b 4 C23Hz~N30iS C, H, S u 99 173-174 17a 1 sir, C ~ Z H ~ ~ ? J ~ O ~ C, H, K B 80 197-2008 lib 4 SH, CdLiN30uS C, H, N D 18 4 r\;HCOCsH&Oz-p 30* 275-284 CaoHmSaOsS C, H, N 19 4 N HCOCeHaNHrp B 70 209-2 12 C ~ O H ~ Z S ~ O ~ S . O . ~C, H ~H,O 1\; 1:ecIystalliLed from AIeOH. b Recrystallized from t-PrOH-Et,O. e Recrystallized from EtOH-EtZO. Recrystallized from PrOH. liecrystallized from EtOH-H,O. Recrvstallized from DNF-EtOH. g Recrystallized from EtOH. Recrystallized from hIeOEtOH4
-
e
Il.
249
3 3 3 4
NHCOCsH,SO,F-p NHCOC~H~PO~F-VL NHCOCH2Br NHCOC6HdSOjF-p XHCOC~H~SO~F-VL NIICONHC~H~YJ~F-VL p-SIlCOC6HqNHCOCsH~S;OzF-p SHCOCH2Br NO1 SO1
100 nil of 9 : l lIeOEtOH-H20 was shaken with Hz at 2-3 atm in the presence of 250 mg of Syo Pd-C for about 90 min when rediiction was complete. The filtered solution was evaporated in oacuo. Recrystallization from EtOH gave 0.82 g (3876) of product, mp 197-200'; an additional 0.30 g (total 80%) was isolated from the filtrate. See Table I1 for additional data. p-(p-Bromoacetamidophenoxybutoxy)benzamidine Benzenea solution of 288 mg (0.5 sulfonate (11). Method C.-To mmole) of 17b in 1 ml of D M F stirred in an ice bath was added a solution of 192 mg (0.7,s mmole) of bromoacetic anhydride in 2 ml of DMF.15 The ice bath was removed and after 1 hr the solution was dilrited with several volumes of EtgO. The solvent. was decanted from the oil and the latter was cryst,allizedby addition of EtOH-EtlO. Recrystallization from EtOH containing 100 mg of C6H5S03.H20and a second recryst,allization from EtOH gave 100 mg (34y0) of white crystals, mp 198-200", that gave a positive 4-(p-riitrobenzyl)pyridinetest for activated halogen.': See Table I1 for additional data. p - [ p-( p-Fluorsulfonylbenzamido)phenoxybutoxy] benzamidine Benzenesulfonate (7). Method D.-To a stirred solution of 228 mg (0.05 mmole) of 17b in 1 ml of D M F was added 0.73 ml of 1 A f Et3K in DMF. Then a solut,ion of 150 mg (0.67 mmole) of p-fluorosulfonylbenzoyl chloride in DXIF was added diopwise over a period of about 5 min, the flask being cooled with a 15-20" HzO bath. After 90 min at ambient temperature, the solution was treated with 17.3 mg (1 mmole) of CsHESOaH.H20 in 8 ml of H,O. The mixture was warmed on a steam bath, then 2-PrOH was added until a clear solution was obtained. The solution was kept a t - 15' for several hours. The gummy solid was recrystallized from 3IeOH; yield 60 mg (18%), mp 280283". See Table I1 for addit)ionaldata. Method E was the same with omission of Et3N. 4-Fluorosulfonyl-4'-methoxybenzanilide(12).-To a stirred solution of 615 mg (5 mmoles) of p-anisidine and 0.51 g (5 mmoles) of Et3X in 10 ml of CHCl, cooled in an ice bath was added dropwise a solution of 1.23 g ( 5 . 5 mmoles) of p-fluorosulfonylbeiizoyl chloride in 10 ml of cIIc13 over a period of 20 min. The prodlict (mp 195-199O) was collected on a filter and washed with CI%c13. Two recrystallizations from CHCls gave 0.35 g (36y0)of analytical sample as white crystals, mp 199-200". Anal. (CiiHi2FN04S) C, H, Tu'. Enzyme Assays.-The reversible inhibition assays were performed with crystalline trypsin from bovine pancreas (Sigma Chemical Co.) and 50 p M ?j-benzoyl-DL-arginine p-nitroanilide (BASA)'s in pH 7.4 Tris buffer containing 10% DMSO as pre-
viously described. The irreversible inhibition assays were performed as follows. The velocity of the enzyme reaction with 50 p X BA4NAwas observed to be proportional to the enzyme concentration. The amount of spont,aneous enzyme inact,ivation in 10% DRISO at pH 7.4 was about 10%/hr; the inhibitor inactivations in Table I were corrected for this thermal inactivation of an enzyme control that was run simultaneously. The buffer employed was 0.05 'it. Tris (pH 7.4); additional buffers employed were 0.05 -If Tris-maleate (pH 6.5), 0.05 hi' Tris (pH 8.4), and 0.05 ilf citrate (pH 5.0). Bulk enzyme m-as dissolved in cold 1 mM HC1 at 0.9 mg/ml and stored at 0-5" BAXA was stored as a 3.1 miJf solut,ion in DMSO in a brown bottle. In two tubes were placed 0.50 ml of enzyme solution and 1.30 ml of 0.05 M buffer of the desired pH. After 5 min in a 37" bath 0.90 ml was withdrawn from one tube, labeled Io,and placed in an ice bath; then 100 pl of a D3ISO solution of inhibitor was added to the remainder in a 3 i " bath, which was labeled I,, and the time was noted. To the second tube serving as a control was added 200 p1 of DMSO, then 1.00 ml was withdrawn, placed in an ice bath, and labeled C1. The tubes labeled 1, and C2 were then heated in the 37" bath for the specified time, then placed in an ice bath unt,il ready for assay. The amount of remaining enzyme in each aliquot, except the Io tube, was assayed as follows. I n a 3-ml glass cuvette were placed 2.83 ml of pH 7.4 Tris buffer (regardless of the incubation pH) and 50 pl of 3.1 m J i BANA. The enzyme reaction was then started by addition of 200 pl of C1 or other aliquot. The increase in optical density a t 410 mp was followed with a Gilford 2000 spectrophotometer: the Cl aliquot gave an optical density change of 0.007-0.012 OD unit/min. The velocities in OD/min, which are proportional to the remaining active enzyme concent,ration, were plotted on a log scale against time on a linear scale. Each aliquot. was run in duplicate, or in triplicate if necessary. Just before assay, the still ice-cold Ioaliquot was treated with 100 pl of the DMSO solution of inhibitor, then ashayed immediately. This procedure is satisfactory for a routine screen for a pliis or miniis answer. With a positive compornid, the size of the inhibitor incubation is increased threefold, the IOaliquot is removed as before, then the remainder is aliquoted a t appropriate t'imes such as 2, 4, 8, 16, and 30 min. (18) B. F. Erlanger, N. Kokowsky, and W. Cohen, Arch. Biochem. Biophzls., 96, 271 (1961).